A Comprehensive Mechanistic Model for

Upward Two-Phase Flow in Wellbores

A.M. Ansari, Pakistan Petroleum Ltd.; N.D. Sylvester, U. of Akron; and C. Sarica, O. Shoham, and
J.P. Brill, U. of Tulsa
Summary. A comprehensive model is formulated to predict the flow behavior for upward two-phase flow. This model is composed of
a model for flow-pattern prediction and a set of independent mechanistic models for predicting such flow characteristics as holdup and
pressure drop in bubble, slug, and annular flow. The comprehensive model is evaluated by using a well data bank made up of 1,712 well
cases covering a wide variety of field data. Model performance is also compared with six commonly used empirical correlations and the
Hasan-Kabir mechanistic model. Overall model performance is in good agreement with the data. In comparison with other methods, the
comprehensive model performed the best.
Introduction
Two-phase flow is commonly encountered in the petroleum, chemi-
cal, and nuclear industries. This frequent occurrence presents the
challenge of understanding, analyzing, and designing two-phase
systems.
Because of the complex nature of two-phase flow, the problem
was first approached through empirical methods. The trend has
shifted recently to the modeling approach. The fundamental postu-
late of the modeling approach is the existence of flow patterns or
flow configurations. Various theories have been developed to pre-
dict flow patterns. Separate models were developed for each flow
pattern to predict flow characteristics like holdup and pressure drop.
By considering basic fluid mechanics, the resulting models can be
applied with more confidence to flow conditions other than those
used for their development. .
Only Ozon et at.
l
and Hasan and Kabir2 published studies on
comprehensive mechanistic modeling of two-phase flow in vertical
pipes. More work is needed to develop models that describe the
physical phenomena more rigorously.
The purpose of this study is to formulate a detailed comprehen-
sive mechanistic model for upward two-phase flow. The compre-
hensive model first predicts the existing flow pattern and then calcu-
lates the flow variables by taking into account the actual
mechanisms of the predicted flow pattern. The model is evaluated
against a wide range of experimental and field data available in the
updated Tulsa U. Fluid Flow Projects (TUFFP) well data bank. The
performance of the model is also compared with six empirical cor-
relations and one mechanistic model used in the field.
Flow-Pattern Prediction
Taitel et al.
3
presented the basic work on mechanistic modeling of
flow-pattern transitions for upward two-phase flow. They identified
four distinct flow patterns (bubble, slug, chum, and annular flow)
and formulated and evaluated the transition boundaries among them
(Fig. 1). Barnea et al.
4
1ater modified the transitions to extend the
applicability of the model to inclined flows. Barnea
5
then combined
flow-pattern prediction models applicable to different inclination
angle ranges into one unified model. Based on these different works,
flow pattern can be predicted by defining transition boundaries
among bubble, slug, and annular flows.
Bubble/Slug Transition. Taitel et al.
3
gave the minimum diameter
at which bubble flow occurs as
For pipes larger than this, the basic transition mechanism for bubble
to slug flow is coalescence of small gas bubbles into large Taylor
bubbles. This was found experimentally to occur at a void fraction
Copyright 1994 Society of Petroleum Engineers
SPE Production & Facilities, May 1994
of about 0.2S. Using this value of void fraction, we can express the
transition in terms of superficial and slip velocities:
VSg = 0.2Svs + 0.333vSL' ......................... (2)
where Vs is the slip or bubble-rise velocity given by6
y.
Vs = 1.53 [ .
(3)
This is shown as Transition A in Fig. 2.
Dispersed Bubble Transition. At high liquid rates, turbulent forces
break large gas bubbles down into small ones, even at void fractions
exceeding 0.2S. This yields the transition to dispersed bubble flow
5
:
( )
0.5
VSg
= O.72S + 4.1S + .
VSg vSL
(4)
This is shown as Transition B in Fig. 2.
At high gas velocities, this transition is governed by the maxi-
mum packing of bubbles to give coalescence. Scott and Kouba
7
concluded that this occurs at a void fraction of 0.76, giving the tran-
sition for no-slip dispersed bubble flow as
VSg = 3.l7v
SL
................................ (S)
This is shown as Transition C in Fig. 2.
Transition to Annular Flow. The transition criterion for annular
flow is based on the gas-phase velocity required to prevent the en-
trained liquid droplets from falling back into the gas stream. This
gives the transition as
(6)
shown as Transition D in Fig. 2.
Barnea
5
modified the same transition by considering the effects
of film thickness on the transition. One effect is that a thick liquid
film bridged the gas core at high liquid rates. The other effect is in-
stability of the liquid film, which causes downward flow of the film
at low liquid rates. The bridging mechanism is governed by the
minimum liquid holdup required to form a liquid slug:
HLF > 0.12, ..................................... (7)
where HLF is the fraction of pipe cross section occupied by the liq-
uid film, assuming no entrainment in the core. The mechanism of
film instability can be expressed in terms of the modified Lockhart
and Martinelli parameters, XM and YM,
143
t
t t t
.
. .
...
.
. .
. ' .
t t t
t
BUBBLE
FCOW
SLUG
FLOW
CHURN
FLOW
ANNULAR
FLOW
Fig. 1-Flow patterns in upward two-phase flow.
DISPERSED BUBBLE
A
BUBBLY
B
SLUG OR CHURN
D
D
0.01 L..... __ -L-'-__ .J......_---'--.L._--'_...l..--..J
0.01 0.1 1 10
SUPERFICIAL GAS VELOCITY (mls)
Fig. 2-Typical flow-pattern map for well bores.
100
_ 2-1.5H
LF
2
YM - HiF(1-1.5H
LF
/
M
' ........ (8)
where X
M
=

(
dP) ,
dL sc
......................... (9)
g sin 8(pcpa
YM = (:ftc , ........................... (10)
and B=( I-FE)2UFIfsL). From geometric considerations, HLF can be
expressed in terms of minimum dimensionless film thickness,
as
HLF = 4Qrnin(l-Qrnin)' .......................... (11)
144
To account for the effect of the liquid entrainment in the gas core,
Eq. 7 is modified here as
( HLfd > 0.12 .......................... (12)
Annular flow exists if VSg is greater than that at the transition giv-
en by Eq. 6 and if the two Barnea criteria are satisfied. To satisfy the
Barnea criteria, Eq. 8 must first be solved implicitly for HLF
is then calculated from Eq. II; ifEq. 12 is not satisfied, annular flow
exists. Eq. 8 can usually be solved for by using a second-order
Newton-Raphson approach. Thus, Eq:-8 can be expressed as
2-1.5H
LF
2
F(Qrnin) = Y M Hil
l
-1.5H
LF
/ M .. (13)
and

+ Hil
l
-1.
5H
LF)2 .................... (14)
The minimum dimensionless film thickness is then determined it-
eratively from
F(Q.rnin )
= . J (15)
-rrun
j
+1 _mlDj . )"
-mInj
A good initial guess is Q.min =0.25.
Flow-Behavior Prediction
After the flow patterns are predicted, the next step is to develop
physical models for the flow behavior in each flow pattern. This step
resulted in separate models for bubble, slug, and annular flow.
Chum flow has not yet been modeled because of its complexity and
is treated as part of slug flow. The models developed for other flow
patterns are discussed below.
Bubble Flow Model. The bubble flow model is based on Caetano's8
work for flow in an annulus. The bubble flow and dispersed bubble
flow regimes are considered separately in developing the model for
the bubble flow pattern.
Because of the uniform distribution of gas bubbles in the liquid
and no slippage between the two phases, dispersed bubble flow can
be approximated as a pseudosingle phase. With this simplification,
the two-phase parameters can be expressed as
PTP = PLAL + pg(I-AL), .......................... (16)
!1-TP = !1-LAL + !1-g(1-AL), ....................... (17)
and VTP = VMV
SL
+ vS
g
' (18)
where AL = vsdv
m
(19)
For bubble flow, the slippage is considered by taking into account
the bubble-rise velocity relative to the mixture velocity. By assum-
ing a turbulent velocity profile for the mixture with the rising bubble
concentrated more at the center than along the pipe wall, we can ex-
press the slippage velocity as
Vs = v
g
-1.2v
m
(20)
Harmathy6 gave an expression for bubble-rise velocity (Eq. 3).
To account for the effect of bubble swarm, Zuber and Hench
9
modi-
fied this expression:
v.
[
gOL(PCPg)] n'
Vs = 1.53 pi H L' .......... (21)
SPE Production & Facilities, May 1994
DEVELOPED
TAYLOR
BUBBLE--H
I
I
VTO
LSU DEVELOPING
TAYLOR
BUBBLE
(0) DEVELOPED SLUG UNIT (b) DEVELOPING SLUG UNIT
Fig. 3-Schematic of slug flow.
where the value of n' varies from one study to another. In the present
study, n'=0.5 was used to give the best results. Thus, Eq. 20 yields
v.
1.53 [ = -1.2VM' ........... (22)
This gives an implicit equation for the actual holdup for bubble
flow. The two-phase flow parameters can now be calculated from
PTP = PLH
L
+ pg(l-HL) .......................... (23)
and !lTP = !lLHL + !lg(1-H
L
) ...................... (24)
The two-phase pressure gradient is made up of three components.
Thus,
(:) = (:). + + ................. (25)
The elevation pressure gradient is given by
(:). = PTpgsin9 . ........................... (26)
The friction component is given by
(:t = fTPP;;V}p, ............................. (27)
where fTP is obtained from a Moody diagram for a Reynolds number
defined by
N
Re
PTt
Tpli
. . ................................ (28)
TP ,.,.TP
Because bubble flow is dominated by a relatively incompressible
liquid phase, there is no significant change in the density of the flow-
ing fluids. This keeps the fluid velocity nearly constant, resulting in
essentially no pressure drop owing to acceleration. Therefore, the
acceleration pressure drop is safely neglected, compared with the
other pressure drop components.
Slug Flow Model. Fernandes et al.
lD
developed the first thorough
physical model for slug flow. Sylvester!! presented a simplified ver-
SPE Production & Facilities, May 1994
sion of this model. The basic simplification was the use of a correla-
tion for slug void fraction. These models used an important assump-
tion of fully developed slug flow. McQuillan and Whalley12
introduced the concept of developing flow during their study of
flow-pattern transitions. Because of the basic difference in flow ge-
ometry, the model treats fully developed and developing flow sepa-
rately.
For a fully developed slug unit (Fig. 3a), the overall gas and liquid
mass balances give
VSg = [3V
g
TB(1-H
LTB
) + (H1)vgLS (I-HLLS) ........... (29)
and v SL = (1- [3)v LLSH LLS - [3 V LTB H LTB, .............. (30)
respectively, where
[3 = LTBI Lsu .................................. (31)
Mass balances for liquid and gas from liquid slug to Taylor bubble
give
(vTB cc-vLLS)H
LLS
= [VTB- (-VLTB)]HLTB ............. (32)
and (VTl,v
g
LS)(1-H
LLS
) = (VTl,VgTB)(l-H
LTB
). . ......... (33)
The Taylor bubble-rise velocity is equal to the centerline velocity
plus the Taylor bubble-rise velocity in a stagnant liquid column; i.e.,
y,
[
gd(PCPg)]
VTB = 1.2vm + 0.35 PL .................. (34)
Similarly, the velocity of the gas bubbles in the liquid slug is
v.
_ [gaL(Pcpg)] ,,0.5
VgLS - 1.2vm + 1.53 Pi. fliLs,' .......... (35)
where the second term on the right side represents the bubble-rise
velocity defined in Eq. 21.
The velocity of the falling film can be correlated with the film
thickness with the Brotz
13
expression,
v
LTB
= jI96.7g1h, ............................ (36)
where b
L
, the constant film thickness for developed flow, can be ex-
pressed in terms of Taylor bubble void fraction to give
v
LTB
= 9.916[gd(l-jH
gTB
)]"" . ..................... (37)
145
The liquid slug void fraction can be obtained by Sylvester'sl1 cor-
relation and from Fernandes et al.'slO and Schmidt's14 data,
VSg
HgLS = 0.425 + 2.65v
m
......................... (38)
Eqs. 29 or 30,31 through 35, 37, and 38 can be solved iteratively
to obtain the following eight unknowns that define the slug flow
model: fl, HLTB, HgLS, VgTB' VLTB, VgLS, VLLS, and VTB. Vo and Sho-
ham
15
showed that these eight equations can be combined algebra-
ically to give
(9.916!id)(1-b-HLTB)0.5HLTn-VTB(l-HLTB) + A = 0,
..................... (39)
Gd(PcPg) 0.5
[
{
[ ]
0.25 }]
X Vm-HgLS l.53 pi (l-
H
gd .
..................... (40)
With VTB and HgLS given by Eqs. 34 and 38, respectively, A can
be readily determined from Eq. 40. Eq. 39 is then used to find HLTB
with an iterative solution method. Defining the left side ofEq. 39 as
F(HLTB), then
F(H
LTB
) = (9.916!id)(I-b-HLTB)o.5HLTn-VTB(1-HLTB) + A.
..................... (41)
Taking the derivative of Eq. 41 with respect to HLTB yields
F'(H
LTB
) = VTB + (9.916!id)
X [(1-
b
-HLTB)o.5 + HLTB ].
4 j(l-H
LTB
)( 1-b-HLTB)
..................... (42)
HLTB, the root of Eq. 39, is then determined iteratively from
F(H
LTB
)
HLTBj+l = HLTBj- ....................... (43)
J
The step-by-step procedure for determining all slug flow vari-
ables is as follows.
1. Ca1culate VTB and HgLS from Eqs. 34 and 38.
2. Using Eqs. 40 through 43, determine HLTB. A good initial
guess is HLTB=0.15.
3. Solve Eq. 37 for VLTB. Note that H
g
TB=I-HLTB.
4. Solve Eq. 32 for VLLS. Note that HLLS=l-HgLS.
5. Solve Eq. 35 for VgLS.
6. Solve Eq. 33 for VgTB.
7. Solve Eq. 29 or 30 for fl.
8. Assuming that LLS=30d, ca1culate Lsu and LTB from the defini-
tion of fl.
To model developing slug flow, as in Fig. 3b, we must determine
the existence of such flow. This requires ca1culating and comparing
the cap length with the total length of a developed Taylor bubble.
The expression for the cap length is 12
2
1 [V
NgTB
Vm ]
Lc = 2 V
TB
+ V-(l-H
NLTB
) --H , .......... (44)
g
where VNgTB and HNLTB are ca1culated at the terminal film thickness
(called Nusselt film thickness) given by
[
1/3
3 VNLTB#L(I-HNLTB)]
ON = 4
d
g(PCpg) ... ................ (45)
146
The geometry of the film flow gives HNLTB in terms of oN as
2
HNLTB = 1-( 1- ............................ (46)
To determine VNgTB, the net flow rate of ON can be used to obtain
(I-HLLS)
VNgTB = vTB-(vTB-vgLS)(1 H ) .................. (47)
- NLTB
The length of the liquid slug can be ca1culated empirically from
LLS = C'd, .................................... (48)
where C' was found
16
to vary from 16 to 45. We use C'=30 in this
study. This gives the Taylor bubble length as
LTB = [LLS/(l-{J)J{J . ............................. (49)
From the comparison of 4: and LTB, if 4: 2: LTB, the flow is devel-
oping slug flow. This requires new values for LrB, HtTB' and
VtTB ca1culated earlier for developed flow.
For LtB, Taylor bubble volume can be used:
LTB
v:'TB = I AiB (L)dL, ............................ (50)
where AiB can be expressed in terms of local holdup hLTB(L),
which in tum can be expressed in terms of velocities by using Eq.
32. This gives
A*(L) = [1 (VTlrVLLS)HLLS]A
TB /2ii p... .............. (51)
The volume can be expressed in terms of flow geometry as
Substitution of Eqs. 51 and 52 into Eq. 50 gives
(
LrB + LLS) LLS
VSg -vgLS(l-H
LLS
)-
VTB VTB
LITB [ (VTB - VLLS)HLLS]
1 /2ii dL . .................... (53)
o
Eq. 53 can be integrated and then simplified to give
*2 (-2ab-4c
2
) * b
2
_
LTB + LTB + - - 0, .................. (54)
a2 a
2
where a=l-vs
g
IVTB, ................................ (55)
VSg-vgd2-HLLS)
b = LLS, ......................... (56)
V
TB
VTB-V
LLS
and c = Iii H
LLS
............................. (57)
After calculating LiB' the other local parameters can be calcu-
lated from
v ITB (L) = /2ii-VTB ............................ (58)
* (VTIrVLLS)HLLS
and hLTB (L) = .fiii ...................... (59)
In calculating pressure gradients, we consider the effect of vary-
ing film thickness and neglect the effect of friction along the Taylor
bubble.
SPE Production & Facilities, May 1994
For developed flow, the elevation component occurring across a
slug unit is given by
(%t = [(l-P)PLS + Ppglg sine, .................. (60)
where PLS = PLH
us
+ pi1-Hus) . .................. (61)
The elevation component for developing slug flow is given by
= [(l-P *)PLS + P * PTBAlg sin e, ............ (62)
where PTBA is based on average void fraction in the Taylor bubble
section with varying film thickness. It is given by
PTBA = PLH
LTBA
+ Pg(l-HLTBA)' .................. (63)
where HLTBA is obtained by integrating Eq. 59 and dividing by Lh,
giving
2(VTS-VLLS)H
us
HLTBA = j2gLfB ........................ (64)
The friction component is the same for both the developed and de-
veloping slug flows because it occurs only across the liquid slug.
This is given as
(
dP) iLSPLSv;"
dL f = .......................... (65)
where P should be replaced by P* for developing flow. ks can be
calculated by using
NRe = PLSvmd//l
LS
............................ (66)
LS
For the pressure gradient due to acceleration, the velocity in the
film must be considered. The liquid in the slug experiences decel-
eration as its upward velocity of VLLS changes to a downward veloc-
ity of VLTB. The same liquid also experiences acceleration when it
exits from the film with a velocity VLTB into an upward moving liq-
uid slug of velocity VLLS. If the two changes in the liquid velocity
occur within the same slug unit, then no net pressure drop due to ac-
celeration exists over that slug unit. This happens when the slug
flow is stable. The correlation used for slug length is based on its
stable length, so the possibility of a net pressure drop due to accel-
eration does not exist. Therefore, no acceleration component of
pressure gradient is considered over a slug unit.
Annular Flow Model. A discussion on the hydrodynamics of annu-
lar flow was presented by Wallis.J7 Along with this, Wallis also
presented the classic correlations for entrainment and interfacial
friction as a function of film thickness. Later, Hewitt and Hall-Tay-
lor
l8
gave a detailed analysis of the mechanisms involved in an an-
nular flow. All the models that followed later are based on this ap-
proach.
A fully developed annular flow is shown in Fig 4. The conserva-
tion of momentum applied separately to the core and the film yields
Ac( t-r;S;-PcACg sin e = 0 .................... (67)
+r;S;-rFSrpLAFgsine = O ........... (68)
The core density, Pc, is a no-slip density because the core is con-
sidered a homogeneous mixture of gas and entrained liquid droplets
flowing at the same velocity. Thus,
Pc = PLALc + Pg(l-ALC)' ......................... (69)
FEv
SL
where ALe = + F ........................... (70)
VSg EVSL
SPE Production & Facilities, May 1994
- I _

--I
LIQUID FILM ----II6-!
I
-I-
ENTRAINED I
LI Q UI D DROPL -,
V
. C
.-
I ._
:t : .. t V
F
8
Fig. 4-Schematic of annular flow.
FE is the fraction of the total liquid entrained in the core, given by
Wallis as
FE = l-exp[-O.l25(vcr;t-1.5)1, .................... (71)
Vsg/l g (pg)V2
where Vcril = 10,000--0- -P .................... .
L L
(72)
The shear stress in the film can be expressed as
v
2
r
F
= iFPL;' .................................. (73)
where iF is obtained from a Moody diagram for a Reynolds number
defined by
PLvFdHF
N
ReF
= ----;i'L' ............................... (74)
and d
HF
= 4Q(l-Q,)d.
This gives
(75)
(76)
.................... (77)
147
Eq. 77 reduces to
d (1-FE)2 fF(dP)
TF = 4 [4Q(I-Q)FfsL dL SL ' ..................... (78)
where the superficial liquid friction pressure gradient is given by
( :) SL = . ............................. (79)
fsL is the friction factor for superficial liquid velocity and can be
obtained from a Moody diagram for a Reynolds number defined by
NRe = PLVSLd/f.lL .............................. (80)
SL
For the shear stress at the interface,
Tj = /; ................................. (81)
where Vc = vse/(1-2Q)2 ............................ (82)
and/;=fscZ, ...................................... (83)
where Z is a correlating factor for interfacial friction and the film
thickness. Based on the performance of the model, the Wallis ex-
pression for Z works well for thin films or high entrainments, where-
as the Whalley and Hewitt
19
expression is good for thick films or
low entrainments. Thus,
Z = 1 for FE> 0.9 ......................... (84)
(
pg)I/3
and Z = 1 + 24 P
g
Q for FE < 0.9. . ............. (85)
Combining Eqs. 81 through 83 yields
Tj = ........................... (86)
The superficial friction pressure gradient in the core is given by
( : ) se = . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. (87)
where fsc is obtained from a Moody diagram for a Reynolds number
defined by
N
Resc
= Pcvscd/f.lsc> .............................. (88)
vse=FEvsL+VS
g
' . (89)
and f.le = f.lLA
Le
+ f.li1-ALe). . ...................... (90)
The pressure gradient for annular flow can be calculated by sub-
stituting the above equations into Eqs. 67 and 68. Thus,
(
dP) Z (dP) .
dL e = (l_2Q)5 dL se + Peg sm f} ............... (91)
+ ................ (92)
The basic unknown in the above equations is the dimensionless
film thickness, Q. An implicit equation for Q can be obtained by
equating Eqs. 91 and 92. This gives
se-(PcPe)g sin f}
(I-FE)2 IF (dP)
64Q3(1-Q)3 fSL dL SL = O. . ..................... (93)
148
To simplify this equation, the dimensionless approach developed
by Alves et al.
2o
is used. This approach defines the following di-
mensionless groups in addition to previously defined modified
Lockhart Martinelli parameters, XM and YM.
2 (dp/dL)e -gPe sin f}
ifJ
e
= (dp/dL)se ........................ (94)
2 _ (dp/dL)F -gPL sin f}
and ifJF - (dp/dLb ....................... (95)
By using the modified Lockhart Martinelli parameters, Eq. 93 re-
duces to
Z X1
YM 4Q(I-Q)[I-4Q(l-Q)V5 + [4Q(I-Q)P = o ......... (96)
The above equations can be solved iteratively to obtain Q. If Eq.
96 is F(Q), then taking the derivative of Eq. 96 with respect to Q
yields
, Z[4(1-2Q)]
F = [4Q(l-Q)J2[1-4Q(l-Q)]25
Z'
4Q(l-Q)[ 1-4Q( l-Q)]25
2.5Z[4(1-2Q)]
3X1[4(1-2Q)]
[4Q(I--QW . . ........... (97)
4Q(I--Q)[I-4Q(1--QW
5
The Newton-Raphson method can be incorporated to determine
Q, the root of Eq. 96. Thus,
F(Q)
Qj+1 = Qr F'(Q/ .............................. (98)
Once Q is known, the dimensionless groups ifJF and ifJc can be ob-
tained from the following form of Eqs. 91 and 92
= (1_;Q)5 .................................. (99)
2 (1-FE)2 fF
and ifJF = [1-(1-2Q)2]2fsL X
{
}
-Y M[l-(1-2Q)2]2 . . ................... (100)
Alves
20
stated that Eq. 100 can be expressed as
ifJ} = . ................................ (101)
The total pressure gradient can then be obtained from either Eq. 94
or 95 because the pressure gradient in the film and core must be the
same. Thus,
(
d
P
) _ (dP) _ 2 (dP) .
dL T - dL e - ifJe dL se + gPesmf} .......... (102)
or (:t = (:t = + gPL sin f} . ......... (103)
Note that the above total pressure gradient equations do not in-
clude accelerational pressure gradient. This is based on results
found by Lopes and Dukler
21
indicating that, except for a limited
range of high liquid flow rates, the accelerational component result-
SPE Production & Facilities, May 1994
TABLE 1-RANGE OF WELL DATA
Nominal Diameter Oil Rate Gas Rate Oil Gravity
Source (in.) (STB/D) (MSCF/D) ("API)
Old TUFFP' Data 1 to 8 Oto 10,150 1.5 to 10,567 9.5to 70.5
Bank
Govier and 2 to 4 8 to 1,600 114 to 27.400 17 to 112
Fogarasi
22
Asheim
23
27/8 to 6 720 to 27,000 740 to 55,700 35 to 86
Chierici et 81.24 27/8 to 5 0.3 to 69 6to 27,914 8.3 to 46
Prudhoe Bay 5'h to 7 600 to 23,000 200 to 110,000 24 to 86
'Includes data from Poettmann and Carpenter.
25
Fancher and Brown.
26
Hagedorn,27 Baxendell and Thomas.
2B
Orkiszewski.
29
Espanol.
30
Messulam.31 Camacho.32 and field data from several oil companies.
ing from the exchange of liquid droplets between the core and the
film is negligible.
Evaluation
The evaluation of the comprehensive model is carried out by
comparing the pressure drop from the model with the measured data
in the updated TUFFP well data bank that comprises 1,712 well
cases with a wide range of data, as given in Table 1. The perfor-
mance of the model is also compared with that of six correlations
and another mechanistic model that are commonly used in the petro-
leum industry.
Criteria for Comparison with Data
The evaluation of the model using the data bank is based on the fol-
lowing statistical parameters.
Average percent error:
El (r2>ri) x 100, ........................ (104)
.=1
where eri = 6.Pic;,.c-6.Pimeas. . ....................... (105)
Pimeas
E\ indicates the overall trend of the performance, relative to the
measured pressure drop.
Absolute average percentage error:
E2 = (k! le
ri
l)xl00. . ........................ (106)
.=1
E2 indicates how large the errors are on the average.
Percent standard deviation:
n
E3 = I
........................ (107)
i=1
E3 indicates the degree of scattering of the errors about their average
value.
Average error:
E4 = (k! e
i
), ................ (l08)
1=1
'where ej = 6.Pjcalc-6.Pimeas' ........................ (109)
E4 indicates the overall trend independent of the measured pressure
drop.
Absolute average error:
E5 = ............................. (110)
E5 is also independent of the measured pressure drop and indicates
the magnitude of the average error.
Standard deviation:
n
E6 = I
(ei-E4)2
--1-' ........................... (111)
n-
j=}
E6 indicates the scattering of the results, independent of the mea-
sured pressure drop.
Criteria for Comparison With Other
Correlations and Models
The correlations and models used for the comparison are a modified
Hagedorn and Brown,27 Duns and Ros,33 Orkiszewski
29
with Trig-
gia correction,34 Beggs and Bri1l
35
with Palmer correction,36 Muk-
herjee and Brill,37 Aziz eta/., 38 and Hasan and Kabir.
2
.
39
The com-
parison is accomplished by comparing the statistical parameters.
The comparison involves the use of a relative performance factor
defined by
TABLE 2-RELATIVE PERFORMANCE FACTORS
EDB VW OW VNH ANH AB AS VS SNH VSNH AAN
n 1712 1086 626 755 1381 29 1052 654 745 387 70
MODEL 0.700 1.121 1.378 0.081 0.000 0.143 1.295 1.461 0.112 0.142 0.000
HAGBR 0.585 0.600 0.919 0.876 0.774 2.029 0.386 0.485 0.457 0.939 0.546
AZIZ 1.312 1.108 2.085 0.803 1.062 0.262 1.798 1.764 1.314 1.486 0.214
DUNRS 1.719 1.678 1.678 1.711 1.792 1.128 2.056 2.028 1.852 2.296 1.213
HASKA 1.940 2.005 2.201 1.836 1.780 0.009 2.575 2.590 2.044 1.998 1.043
BEGBR 2.982 2.908 3.445 3.321 3.414 2.828 2.883 2.595 3.261 3.282 1.972
ORKIS 4.284 5.273 2.322 5.838 4.688 1.226 3.128 3.318 3.551 4.403 6.000
MUKBR 4.883 4.647 6.000 3.909 4.601 4.463 5.343 5.140 4.977 4.683 1.516
EBD=entire databank; VW=vertical well cases; DW=deviated well cases; VNH=vertical well cases without Hagedorn and Brown data; ANH=all well cases without Hagedorn and
Brown data; AB=all well cases with 75% bubble flow; AS=all well cases 100% slug flow; VS=vertical well cases with 100% slug flow; SNH=all well cases with 100% slug flow
without Hagedorn and Brown data; VSNH=vertical well cases with 100% slug flow without Hagedorn and Brown data; AAN=all well cases with 100% annular flow; HAGBR=
Hagedorn and Brown correlation; AZIZ=Aziz et al. correlation; DUNRS=Duns and Ros correlation; HASKA=Hasan and Kabir mechanistic model; BEGBR=Beggs and Brill correlation;
ORKIS=Orkiszewski correlation; MUKBR=Mukherjee and Brill correlation.
SPE Production & Facilities. May 1994 149
E5-E5 . E6-E6 .
+ E _Em," + E _E
nun
. . (112)
5
max
5
min
6
max
6
min
The minimum and maximum possible values for Frp are 0 and 6,
indicating the best and worst performances, respectively.
The evaluation of the model in terms of Frp is given in Table 2,
with the best value for each column being boldfaced.
Overall Evaluation. The o v ~ r l l evaluation involves the entire
comprehensive model so as to study the combined performance of
all the independent flow pattern behavior models together. The eval-
uation is first performed by using the entire data bank, resulting in
Col. I of Table 2. Model performance is also checked for vertical
well cases only, resulting in Col. 2 of Table 2, and for deviated well
cases only, resulting in Col. 3 of Table 2. To make the comparison
unbiased with respect to the correlations, a second database was
created that excluded 331 sets of data from the Hagedorn and Brown
study. For this reduced data bank, the results for all vertical well
cases are shown in Col. 4 of Table 2, and the results for combined
vertical and deviated well cases are shown in Col. 5 of Table 2.
Evaluation ofIndividual Flow Pattern Models. The performance
of individual flow pattern models is based on sets of data that are
dominant in one particular flow pattern, as predicted by the transi-
tions described earlier. For the bubble flow model, well cases with
bubble flow existing for more than 75% of the well length are con-
sidered in order to have an adequate number of cases. These results
are shown in Col. 6 of Table 2. Cols. 7 through 10 of Table 2 give
results for well cases predicted to have slug flow exist for 100% of
the well length. The cases used for Col. 7 and 8 were selected from
the entire data bank, whereas the cases used for Cols. 9 and 10 and
11 were selected from the reduced data bank that eliminated the
Hagedorn and Brown data, which is one-third of all the vertical well
cases. Finally, Col. 11 of Table 2 gives results for those cases in the
total data bank that were predicted to be in annular flow for 100%
of the well length.
Complete performance results of each model or correlation
against individual statistical parameters (E1, E6) are given in the
supplement to this paper.
40
Conclusions
From Cols. 1 through 11 of Table 2, the performance ofthe model
and other empirical correlations indicates the following.
1. The overall performance of the comprehensive model is supe-
rior to all other methods considered. However, the overall perform-
ances of the Hagedorn and Brown, Aziz et ai., Duns and Ros, and
Hasan and Kabir models are comparable to that of the model. For
the last three, this can be attributed to the use of flow mechanisms
in these methods. The excellent performance of the Hagedorn and
Brown correlation can be explained only by the extensive data used
in its development and modifications made to the correlation. In
fact, when the data without Hagedorn and Brown well cases are con-
sidered, the model performed the best (Cols. 4 and 5).
2. Although the Hagedorn and Brown correlation performed bet-
ter than the other correlations and models for deviated wells, none
of the methods gave satisfactory results (Col. 3).
3. Only 29 well cases were found with over 75% of the well length
predicted to be in bubble flow. The model performed second best to
the Hasan and Kabir mechanistic model for bubble flow (Col. 6).
4. The performance of the slug flow model is exceeded by the
Hagedorn and Brown correlation when the Hagedorn and Brown
data are included in the data bank (Cols. 7 and 8). The model per-
formed best when Hagedorn and Brown data are not included for all
well cases and all vertical well cases (Cols. 9 and 10).
5. The performance of the annular flow models is significantly
better than all other methods (Col. 11).
150
6. Several variables in the mechanistic model, such as bubble rise
velocities and film thickness, are dependent on pipe inclination
angle. Modifications to include inclination angle effects on these
variables should further improve model performance.
Acknowledgments
We thank the TUFFP member companies whose membership fees
were used to fund part of this research project, and Pakistan Petro-
leum Ltd. for the financial support provided A.M. Ansari.
References
I. Ozon, P.M., Ferschneider, G., and Chwetzoff, A.: "A New Multiphase
Flow Model Predicts Pressure and Temperature Profiles," paper SPE
16535 presented at the 1987 SPE Offshore Europe Conference, Aber-
deen, Sept. 8-11, 1987.
2. Hasan, A.R. and Kabir, C.S.: "A Study of Multiphase Flow Behavior in
Vertical Wells," SPEPE (May 1988) 263.
3. Taitel, Y., Barnea, D., and Dukler, A.E.: "Modelling Flow Pattern Tran-
sitions for Steady Upward Gas-Liquid Flow in Vertical Tubes," AIChE
J. (1980) 26, 345.
4. Barnea, D., Shoham, 0 .. and Taitel, Y.: "Flow Pattern Transition for Ver-
tical Downward Two-Phase Flow," Chern. Eng. Sci. (1982) 37, 741.
5. Barnea, D.: "A Unified Model for Predicting Flow-Pattern Transition
for the Whole Range of Pipe Inclinations," IntI. J. Multiphase Flow
(1987) 13, I.
6. Hannathy, T.Z.: "Velocity of Large Drops and Bubbles in Media oflnfi-
nite or Restricted Extent," AIChE J. (1960) 6, 281.
7. Scott, S.L. and Kouba, G.E.: "Advances in Slug Flow Characterization
for Horizontal and Slightly Inclined Pipelines," paper SPE 20628 pre-
sented at the 1990 SPE Annual Technical Conference and Exhibition,
New Orleans, Sept. 23-26.
8. Caetano, E.F.: "Upward Vertical Two-Phase Flow Through an Annu-
lus," PhD dissertation, U. of Tulsa, Tulsa, OK (1985).
9. Zuber, N. and Hench, J.: "Steady State and Transient Void Fraction of
Bubbling Systems and Their Operating Limits. Part I: Steady State Op-
eration," General Electric Report 62GLloo (1962).
10. Fernandes, R.C., Semait, T., and Dukler, A.E.: "Hydrodynamic Model
for Gas-Liquid Slug Flow in Vertical Tubes," AIChE J. (1986) 29, 981.
11. Sylvester, N.D.: "A Mechanistic Model for Two-Phase Vertical Slug
Flow in Pipes," ASME J. Energy Resources Tech. (1987) 109, 206.
12. McQuillan, K.W. and Whalley, P.B.: "Flow Patterns in Vertical Two-
Phase Flow," IntI. J. Multiphase Flow (1985) 11, 161.
13. Brotz, W.: "Uber die Vorausberechnung der Absorptionsgesch-windig-
keit von Gasen in Stromenden Flussigkeitsschichten," Chern. Ing. Tech.
(1954) 26, 470.
14. Schmidt, Z.: Experimental Study of Gas-Liquid Flow in a Pipeline-Riser
System, MS thesis, U. of Tulsa, Tulsa (1976).
15. Vo, D.T. and Shoham, 0.: "A Note on the Existence of a Solution for
Two-Phase Slug Flow in Vertical Pipes," ASME J. Energy Resources
Tech. (1989) 111,64.
16. Dukler, A.E., Maron, D.M., and Brauner, N.: "A Physical Model for Pre-
dicting the Minimum Stable Slug Length," Chern. Eng. Sci. (1985)
1379.
17. Wallis, G.B.: One-Dimensional Two-Phase Flow, McGraw-Hill Book
Co. Inc., New York City (1969).
18. Hewitt, G.P. and Hall-Taylor, N.S.: Annular Two-Phase Flow. Perga-
mon Press, Houston (1970).
19. Whalley, P.B. and Hewitt, G.F.: "The Correlation of Liquid Entrainment
Fraction and Entrainment Rate in Annular Two-Phase Flow," UKAEA
Report AERE-R9187, Harwell (1978).
20. Alves,I.N. et al.: "Modeling Annular Flow Behavior for Gas Wells," pa-
per presented at the 1988 Winter Annual Meeting of ASME, Chicago,
Nov. 27-Dec. 2.
21. Lopes, J .C.B. and Dukler, A.E.: "Droplet Entrainment in Vertical Annu-
lar Flow and Its Contribution to Momentum Transfer," AIChE J. (1986)
1500.
22. Govier, G.w. and Fogarasi, M.: "Pressure Drop in Wells Producing Gas
and Condensate," J. Cdn. Pet. Tech. (Oct.-Dec. 1975) 28.
23. Asheim. H.: "MONA, An Accurate Two-Phase Well Flow Model Based
on Phase Slippage," SPEPE (May 1986) 221.
24. Chierici, G.L., Ciucci, G.M., and Sclocchi, G.: "Two-Phase Vertical
Flow in Oil Wells-Prediction of Pressure Drop," JPT(Aug. 1974) 927.
25. Poettmann, P.H. and Carpenter, P.G.: "The Multiphase Flow of Gas, Oil
and Water Through Vertical Flow Strings with Application to the Design
and Gas-Lift Installations." Drill. & Prod. Prac . API, Dallas (1952)
257.
SPE Production & Facilities, May 1994
26. Fancher, G.H., and Brown, K.E.: "Prediction of Pressure Gradients for
Multiphase Flow in Tubing," Trans., AIME (1963) 228, 59.
27. Hagedorn, A.R.: "Experimental Study of Pressure Gradients Occurring
during Continuous Two-Phase Flow in Small Diameter Vertical Con-
duits," PhD dissertation, U. of Texas, Austin (1964).
28. Baxendell, P.B.: "The Calculation of Pressure Gradients in High Rate
Flowing Wells," JPT(Oct. 1961) 1023.
29.0rkiszewski, J.: "Predicting Two-Phase Pressure Drops in Vertical
Pipes," JPT(June 1967) 829.
30. Espanol, H.J.H.: "Comparison of Three Methods for Calculating a Pres-
sure Traverse in Vertical Multi-Phase Flow," MS thesis, U. of Tulsa, Tul-
sa, OK (1968).
31. Messulam, S.A.G.: "Comparison of Correlations for Predicting Multi-
phase Flowing Pressure Losses in Vertical Pipes," MS thesis, U. of Tul-
sa, Tulsa, OK (1970)
32. Camacho, C.A.: "Comparison of Correlations for Predicting Pressure
Losses in High Gas-Liquid Ratio Vertical Wells," MS thesis, U. ofTulsa,
Tulsa, OK (1970).
33. Duns, H. Jr. and Ros, N.C.J.: "Vertical Flow of Gas and Liquid Mixtures
in Wells," Proc., 6th World Pet. Congo (1963) 451.
34. Brill, J.P.: "Discontinuities in the Orkiszewski Correlation for predict-
ing pressure Gradients in Wells," J. Energy Res. Tech. (March, 1989) 41,
34.
35. Beggs, H.D. and Brill, J.P.: "A Study of Two-Phase Flow in Inclined
Pipes," JPT(May 1973) 607.
36. Palmer, C.M.: "Evaluation of Inclined Pipe Two-Phase Liquid Holdup
Correlations Using Experimental Data," MS thesis, U. of Tulsa, Tulsa,
OK (1975).
37. Mukherjee, H. and Brill, J.P.: "Pressure Drop Correlations for Inclined
Two-Phase Flow," J. Energy Res. Tech. (Dec. 1985).
38. Aziz, K., Govier, G.W., and Fogarasi, M.: "Pressure Drop in Wells Pro-
ducing Oil and Gas," J. Cdn. Pet. Tech. (July-Sept. 1972) 38.
39. Kabir, C.S. and Hasan, A.R.: "Performance of a Two-Phase GaslLiquid
Flow Model in Vertical Wells," JPSE (1990) 4, 273.
40. Ansari, A.M. et at.: "Supplement to paper SPE 20630, A Comprehen-
sive Mechanistic Model for Upward Two-Phase Flow in Wellbores," pa-
per SPE 28671 available at SPE headquarters, Richardson, TX.
Nomenclature
a = coefficient defined in Eq. 55
A = cross-sectional area of pipe, L, m
2
b = coefficient defined in Eq. 56
c = coefficient defined in Eq. 57
C = constant factor relating friction factor to Reynolds
number for smooth pipes
C' = coefficient defined in Eq. 48
d = pipe diameter, L, m
e = error function
El = average percentage error, %
E2 = absolute average percentage error, %
E3 = standard deviation, %
E4 = average error, mlLt
2
, psi
E5 = absolute average error, mlLt
2
, psi
E6 = standard deviation, mlLt
2
, psi
f = friction factor
FE = fraction of liquid entrained in gas core
Frp = relative performance factor, defined in Eq. 112
g = gravity acceleration, mls
2
h = local holdup fraction
H = average holdup fraction
L = length along the pipe, m
n = number of well cases
n' = exponent to account for the swarm effect on bubble
rise velocity
SPE Production & Facilities, May 1994
NRe = Reynolds number
p = pressure, mlLt
2
, psi
q = flow rate, L3/t m
3
/s
S = wetted perimeter, L, m
v = velocity, Lit mls
V = volume, L3, m
3
X = Lockhart and Martinelli parameter
Y = Lockhart and Martinelli parameter
Z = empirical factor defining interfacial friction
f3 = length ratio, defined in Eq. 31
<5 = film thickness, L, m
<5 = ratio of film thickness to diameter
K = difference
e = absolute pipe roughness, L, m
o = angle from horizontal, rad or deg
A = no-slip holdup fraction
# = dynamic viscosity, kg/m's, kg/m-
s
v = kinematic viscosity, L2/t, m
2
/sq
p = density, mlL
3
, kg/m3
a = surface tension, mlt
2
, dyne/cm
T = shear stress, mlLt
2
, N/m
3
tP = dimensionless groups defined in Eqs. 94 and 95
Subscripts
a = acceleration
A = average
c = Taylor bubble cap, core
crit = critical
e = elevation
f= friction
F= film
g = gas
H = hydraulic
i = ith element
I = interfacial
L = liquid
LS = liquid slug
m = mixture
M = modified
max = maximum
min = minimum
N= Nusselt
p = pipe
r = relative
s = slip
S = superficial
SU = slug unit
t = total
TB = Taylor bubble
TP = two-phase
Superscript
* = developing slug flow
SI Metric Conversion Factor
in. X 2.54* E+OO= cm
'Conversion factor is exact. SPEPF
Original SPE manuscript received for review Sept. 2, 1990. Revised manuscript received
Sept. 29,1993. Paper accepted for publication Dec. 6, 1993. Paper (SPE 20630) first pre-
sented at the 1990 SPE Annual Technical Conference & Exhibition held in New Orleans,
Sept. 23-26.
151
A.M. Ansari. a senior petroleum engineer, joined Pakistan Petro-
leum Ltd. in 1989. He holds a BS degree in mechanical engineer-
ing from NED U., Karachi, and an MS degree in petroleum engi-
neering from the U. of Tulsa. Niehola.s D. Sylvester is dean of
engineering at the U. of Akron. He holds a BS degree from Ohio
U. and a PhD degree from Carnegie-Mellon U., both in chemical
engineering. Previously he served as dean of engineering and
applied sciences and chairman of petroleum engineering at the
U. of Tulsa. Sylvester has numerous publications in the areas of
two-phase and non-Newtonian fluid flow as well as oil field pollu-
tion control. Cern Sa rica is a research associate in the petroleum
engineering department at the U. of Tulsa. He holds BS and MS
degrees in petroleum engineering from Istanbul Technical U.,
and a PhD degree in petroleum engineering from the U. of Tulsa.
He was an assistant professor of petroleum engineering at Istan-
bul Technical U. before joining the U. of Tulsa. His research inter-
ests are multiphase flow in pipes, oil and gas production, and
fluid flow in porous media. Ovadia Shoham is an associate pro-
fessor of petroleum engineering at the U. of Tulsa. He received
his PhD degree in mechanical engineering from Tel Aviv U., his
MS degree in chemical engineering from the U. of Houston, and
his BS degree in chemical engineering from the Technion in Is-
rael. Shoham currently teaches and conducts research in the
area of modeling two-phase flow in pipes and its application in
011 and gas production, transportation, and separation. He has
over fifty publications in the areas of two-phase flow and produc-
tion operations. Shoham served in several capacities in the Tulsa
U. Fluid Flow Projects (1982-1991), including Director of Re-
search. He was a member of the Production Operation Technical
Committee (1990-1992) and served as a member ofthe planning
committee of the Forum Series on Multiphase Flow in 1992.
James P. Brill is the F.M. Stevenson professor of petroleum engi-
neering at the U. of Tulsa and founder and Executive Director of
the Tulsa U. Fluid Flow Projects. He holds a BS degree from the
U. of Minnesota and a PhD in petroleum engineering from the U.
of Texas at Austin. He currently teaches and conducts research
in the areas of two-phase flow in pipes and production design
and has published extensively in these areas. Brill was a member
of the Distinguished Lecturer Committee, a 1982 Distinguished
Lecturer, 1978-81 Engineering Manpower Committee member, a
member of the Education and Professionalism Committee for the
1971 and 1976 Annual Meetings, and a member of the Education
and Accreditation Committee during 1967-70. He has served as
an SPE representative to the ABET Engineering Accreditation
Commission and ex-officio member of the Education and Ac-
creditation Committee since 1990.
152
Ansari
Brill
Sylvester Shoham
SPE Production & Facilities, May 1994